Willi and others,
I am very supportive of projects to improve laboratory comparability of
isotopic data, but I was most surprised to see my name on the list of
people that had "expressed an interest to participate" in this program.
As far as I know, I had not volunteered for this program. Carol Kendall
and I have had several discussions about your proposed program, and we
have some suggestions for it.
In order to gain the most information from your inter-laboratory
comparison, we strongly suggest that you distribute hydrogen gases that
are tied to the absolute hydrogen isotope-ratio scale. This can be done
by relating them to VSMOW water and SLAP water, both of which have been
measured on an absolute basis for hydrogen. There are three absolute
isotope-ratio measurement of these reference materials (using isotope
mixtures) that give the following:
Hagemann et al. (1970, Tellus, v. 22, p. 712) -428.5+/-0.4 permil
De Wit et al. (1980, Geostand. Newsl., v. 4, p. 33) -427.8+/-0.5 permil
Tse et al. (1980, Anal. Chem., v. 52, p. 2445) -428.8+/-1.3 permil
We suggest that a few laboratories be asked to calibrate the gas samples
using VSMOW and SLAP prepared with uranium, prior to general distribution
to investigate mass spectrometer performance. We recommend uranium be
used and zinc avoided because zinc seldom converts hydrogen quantitatively
and the Pyrex combustion tubes commonly used exchange with the hydrogen
(Kendall and Coplen, 1985, Anal. Chem., v. 57, p. 1437). As a
consequence, zinc reaction techniques commonly produce hydrogen isotopic
fractionations as large as 15 permil; often the users are completely
unaware of this because they analyze all their water samples the same way
and the isotopic fractionation can be constant. We suspect that much of
the "apparent" variability in the size of the permil scale in different
laboratories may actually be caused in part by such fractionation.
Hence, zinc should be avoided in such standards work unless one can
demonstrate it gives the same results as uranium. We went to considerable
effort designing a zinc method that does not significantly fractionate
hydrogen isotopes (Kendall and Coplen, 1985). Participants should also
avoid concentrating hydrogen gas on charcoal or uranium, both of which can
cause sizable isotopic fractionations. We suggest that two of the gas
samples be about 0 permil and about -428 permil.
We suggest that a third gas be much lighter. You have proposed a range of
delta values from -400 to +200 permil relative to VSMOW. However, it
seems reasonable that you would certainly want to include the range
encountered in natural terrestrial materials. Therefore, the range should
be expanded to about -836 permil. This would also then include most of
the gaseous hydrogen samples from laboratories that analyze water by
equilibration with gaseous hydrogen. There are numerous laboratories
analyzing water by equilibration, so it may be important. Such samples
might be as negative as -860 permil (Antarctic ice).
A corollary to the above absolute isotope-ratio discussion is that SLAP
must be -428 permil relative to VSMOW. Any laboratory that does not find
-428 has not correctly taken into account one or several mass
spectrometric corrections (memory, H3+, tailing, etc.) or has fractionated
the gas during preparation or handling. Because of the complexity of the
problem, it is far simpler to just normalize to a VSMOW/SLAP scale to -428
permil and be done with it. This, of course, assumes that you are
converting VSMOW water and SLAP water quantitatively into hydrogen gas,
such as with a uranium furnace. If only zinc is used, it is impossible to
separate mass spectrometric corrections from sample preparation errors.
Please note that above we use VSMOW/SLAP scale rather than SMOW/SLAP
scale. This is because Coplen and Clayton (1973, Geochimica, v. 37, p.
2347), using a cycloidal isotope-ratio mass spectrometer capable of
resolving the H1H2+ ion beam from the H3+ ion beam, showed that VSMOW
water is depleted in H-2 by about 1 permil relative to SMOW as defined by
Craig (1961, Science, v. 133, p. 1833).
As I recall, the IAEA recommends that laboratories recalibrate their
laboratories using fresh reference materials every three years or so.
Thus, you may want to provide VSMOW and SLAP to laboratories involved in
your project, or ask them to purchase these reference materials prior to
analyzing your gaseous reference samples.
Producing more than 40 isotopically homogeneous reference gases is not a
trivial matter, as Carol Kendall can tell you--she has produced or
directed the production and calibration of more than ten thousand gaseous
isotopic reference samples. These gases were prepared in batches of 400
each, on an improved version of the vacuum line used to prepare the carbon
dioxide reference materials, NBS 16 and NBS 17. Usually the first time
one prepares a batch of reference materials, they are not isotopically
homogeneous. Generally, about 10 percent of the samples should be
analyzed prior to use in order to verify isotopic homogeneity. Messer
Griesheim in Duisburg has no track record in producing homogeneous gas
reference materials as far as we know. You might want to indicate to
participants which credible isotope laboratory will certify that these
isotopic reference material are homogeneous before they are distributed.
Storage in vessels with valves or stopcocks has been problematic in the
past. To avoid this problem, we have been sealing gaseous hydrogen
reference samples in Pyrex 9-mm tubes for more than a decade; this method
of storage and shipping has been relatively trouble free. Only a tube
breaker is needed for sampling. Once analyzed, the sample should be
discarded, not stored. Stored samples are more often than not
isotopically fractionated.
In case it is of use, we have available for immediate distribution the
following sets of homogeneous gases:
80 9-mm tubes of gaseous hydrogen with delta H-2 near zero
150 9-mm tubes of gaseous hydrogen with delta H-2 of about -125 permil
80 9-mm tubes of gaseous hydrogen with delta H-2 of about -750 permil
Tyler B. Coplen Phone: +1 703 648 5862
U.S. Geological Survey Fax: +1 703 648 5274
431 National Center Email: [log in to unmask]
Reston, Virginia 22092
|
